North Atlantic Rifted Margin Crustal Thickness and Crustal Thinning from
Satellite Gravity Data
N.W. Hurst1, N.J. Kusznir1 & R.S. White2
1
Department of Earth Sciences, Liverpool University, Liverpool, L69 3BX, UK
2
Bullard Laboratories, Cambridge University, Madingley Rd, Cambridge, CB3 0EZ,
UK
Crustal thickness has been derived from satellite gravity data for the North Atlantic
between 50 and 70 degrees N in order to determine rifted margin crustal thinning.
Satellite derived gravity anomaly data (Smith & Sandwell 1997) and bathymetry data
(Gebco 2003) are used to derive the mantle residual gravity anomaly which is then
inverted to give Moho depth. The gravity anomaly inversion to determine Moho depth
is carried out in 3D in the spectral domain. Oceanic lithosphere and stretched
continental margin lithosphere contain a large negative residual thermal gravity
anomaly (up to – 380 mgal) which must be corrected for in order to determine Moho
depth. This thermal gravity correction may be determined for oceanic lithosphere
using oceanic isochron data, and for the thinned continental margin lithosphere using
margin rift age and beta stretching estimates iteratively derived from crustal basement
thickness determined from the gravity inversion. Beta stretching estimates and the
thermal correction rapidly converge within a few iterations. The use of the thermal
correction within the gravity inversion has been tested by producing profiles of
derived crustal thickness running across the mid-Atlantic Ridge from Hatton and
Edoras Banks to their East Greenland conjugate margins. The gravity inversion using
the thermal gravity correction predicts oceanic crustal thicknesses consistent with
seismic observations, while that without the thermal correction predicts much too
great oceanic crustal thicknesses. Because of errors in the location of the oceancontinent transition within the oceanic isochron data set, two approaches may be used
to define the thermal gravity correction for the determination of rifted margin crustal
thickness. In option 1, oceanic isochron data is used to define the location of the COB,
the oceanic lithosphere beta stretching factor is assumed to be infinity, and ocean
isochron data is used to define oceanic age for the calculation of the thermal gravity
anomaly correction. Option 1 has the advantage that it correctly predicts increasing
thermal gravity anomaly correction towards the ocean ridge, but the disadvantage that
it unrealistically assumes that oceanic isochrons (and therefore the COB) are
accurately known. In option 2, the oceanic isochron data is ignored and is not used to
define the COB, and the thermal gravity correction is determined in a similar way to
that for continental lithosphere by using the rift age specified for continental breakup
and lithosphere beta stretching factors derived from crustal basement thickness from
gravity inversion. Option 2 has the advantage that it does not assume the location of
the COB, but the disadvantage that it fails to predict increasing thermal gravity
correction towards the ocean ridge, and as a consequence incorrectly predicts
thickening of oceanic crust with decreasing oceanic age. In the absence of reliable
sediment thickness data, sediment thickness is assumed to be zero giving an upper
bound of Moho depth, crustal basement thickness and continental beta stretching
factor. In the absence of seismic estimates of volcanic addition thickness (magmatic
underplating) for continental margin crust, volcanic addition is assumed to be zero
also giving a lower bound of continental beta stretching factor.